My thoughts, try not to look at too many things at once. In fact, when I’m creating a dashboard or specifically digging for insights I try to follow a one-to-many or many-to-one design, then start slicing the data. Otherwise I’ll tend suffer from paralysis by analysis and start falling down the rabbit hole of causal loops. I try to isolate an event, time frame, or attribute, and look for outliers, trends, relationships. Sometimes less really IS more.
Kasper highlights some really good points in his post below:
http://www.powerpivotblog.nl/dashboard-design-best-practice
And if you’re really bored you can watch this hour long presentation about dashboard design. Unfortunately not as engaging as your average TED talk, but definitely filled with solid advice.

